End of training

README.md

@@ -18,11 +18,11 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [AnonymousCS/populism_xlmr_large](https://huggingface.co/AnonymousCS/populism_xlmr_large) on the None dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.6899
-- Accuracy: 0.1263
-- 1-f1: 0.2242
-- 1-recall: 1.0
-- 1-precision: 0.1263
+- Loss: 0.3888
+- Accuracy: 0.8737
+- 1-f1: 0.0
+- 1-recall: 0.0
+- 1-precision: 0.0
 - Balanced Acc: 0.5
 
 ## Model description
@@ -42,9 +42,9 @@ More information needed
 ### Training hyperparameters
 
 The following hyperparameters were used during training:
-- learning_rate: 1e-05
-- train_batch_size: 64
-- eval_batch_size: 64
+- learning_rate: 3e-05
+- train_batch_size: 16
+- eval_batch_size: 16
 - seed: 42
 - optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
 - lr_scheduler_type: linear
@@ -53,16 +53,14 @@ The following hyperparameters were used during training:
 
 ### Training results
 
-| Training Loss | Epoch | Step | Validation Loss | Accuracy | 1-f1   | 1-recall | 1-precision | Balanced Acc |
-|:-------------:|:-----:|:----:|:---------------:|:--------:|:------:|:--------:|:-----------:|:------------:|
-| 0.7092        | 1.0   | 13   | 0.6915          | 0.1263   | 0.2242 | 1.0      | 0.1263      | 0.5          |
-| 0.6376        | 2.0   | 26   | 0.6960          | 0.1263   | 0.2242 | 1.0      | 0.1263      | 0.5          |
-| 0.6701        | 3.0   | 39   | 0.6895          | 0.1263   | 0.2242 | 1.0      | 0.1263      | 0.5          |
-| 0.665         | 4.0   | 52   | 0.6903          | 0.1263   | 0.2242 | 1.0      | 0.1263      | 0.5          |
-| 0.6719        | 5.0   | 65   | 0.6894          | 0.1263   | 0.2242 | 1.0      | 0.1263      | 0.5          |
-| 0.6665        | 6.0   | 78   | 0.6885          | 0.1263   | 0.2242 | 1.0      | 0.1263      | 0.5          |
-| 0.6505        | 7.0   | 91   | 0.6897          | 0.1263   | 0.2242 | 1.0      | 0.1263      | 0.5          |
-| 0.7155        | 8.0   | 104  | 0.6899          | 0.1263   | 0.2242 | 1.0      | 0.1263      | 0.5          |
+| Training Loss | Epoch | Step | Validation Loss | Accuracy | 1-f1 | 1-recall | 1-precision | Balanced Acc |
+|:-------------:|:-----:|:----:|:---------------:|:--------:|:----:|:--------:|:-----------:|:------------:|
+| 0.5364        | 1.0   | 50   | 0.4089          | 0.8737   | 0.0  | 0.0      | 0.0         | 0.5          |
+| 0.7823        | 2.0   | 100  | 0.3801          | 0.8737   | 0.0  | 0.0      | 0.0         | 0.5          |
+| 0.2821        | 3.0   | 150  | 0.3801          | 0.8737   | 0.0  | 0.0      | 0.0         | 0.5          |
+| 0.4428        | 4.0   | 200  | 0.3835          | 0.8737   | 0.0  | 0.0      | 0.0         | 0.5          |
+| 0.4079        | 5.0   | 250  | 0.3823          | 0.8737   | 0.0  | 0.0      | 0.0         | 0.5          |
+| 0.4125        | 6.0   | 300  | 0.3888          | 0.8737   | 0.0  | 0.0      | 0.0         | 0.5          |
 
 
 ### Framework versions

config.json

@@ -3,12 +3,12 @@
   "architectures": [
     "XLMRobertaForSequenceClassification"
   ],
-  "attention_probs_dropout_prob": 0.1,
+  "attention_probs_dropout_prob": 0.2,
   "bos_token_id": 0,
-  "classifier_dropout": null,
+  "classifier_dropout": 0.2,
   "eos_token_id": 2,
   "hidden_act": "gelu",
-  "hidden_dropout_prob": 0.1,
+  "hidden_dropout_prob": 0.2,
   "hidden_size": 1024,
   "initializer_range": 0.02,
   "intermediate_size": 4096,

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5f064306f763f6608967e2eb1ad904ec40d6b4b7556ca654f0a37de950774732
+oid sha256:d783e5d27b12c4ba6a88dafcb54f4155ab0126cbd1e28241737c10ee37156841
 size 2239618672

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:847983e481b8926cf911846227209b62e677b5759f3b35f4caffc477560e89c9
+oid sha256:31f67fc2ad51264e7d3ef066240c0365fd2610644940e5ebdea44c9f2185d015
 size 5432